THE POSITION OF SPONGES, ETC. 445 



transformed into flattened cells. F'or the present I think, 

 however, we are justified in regarding the notion as one of 

 the many assumptions which, founded on a mistaken theo- 

 retical basis, and opposed rather than supported by the 

 facts, have dominated and falsified sponge morphology. 

 The physiology of sponges, so far as it is known, does 

 not present much analogy with the Metazoa generally, or with 

 CQijlenterates in particular. Although much is uncertain or 

 unknown, three things are fairly established. First, that 

 the osculum is always an exhalant opening, fi-om which 

 issue forth the currents which enter at the pores and stream 

 along the canals. Secondly, that ingestion may be per- 

 formed by either layer. Numerous experiments by Lieber- 

 kiihn, Metschnikoff and others, have shown beyond all 

 doubt the capacity of the collar cells to take up small food 

 particles from the water which apparently they transfer to 

 the amoeboid cells. On the other hand Metschnikoff (1879) 

 showed clearly that in some sponges iyReiiiera) the food 

 particles are ingested by cells in the dermal layer, and his 

 observations have been confirmed and extended by sub- 

 sequent investigations.^ Metschnikoff also confirmed the 

 observation of Lieberkiihn (1857, p. 388) to the effect that 

 in Spongi//a, Infusoria carried in by the current through the 

 pores into the subdermal spaces and inhalant canals were 

 absorbed there by the wandering cells ; a fact which helps to 

 explain the great development of the inhalant canal system 

 in higher sponges, and throws light on the superior benefit 

 derived by the sponge from the more complicated canal 

 systems as compared with simpler forms such as the Ascons, 

 which are probably capable only of ingesting microscopic 

 particles by means of their collar cells. A third principle 

 of sponge physiology is that all digestion is intracellular, 

 and apparently performed by the wandering amcjebocytes, 

 the " cellules digestives pigment^es " of Topsent. So far, 

 therefore, as conclusions can be drawn from the physiology, 



1 E.g., by Lendenfeld in Aplysilla violacea (uber Coelenteralen der 

 Siidsee II., etc., Zeitschr.f. wiss. Zool. xxxviii., 1883, p. 252); but in his 

 more recent work {ib., xlviii., 1889, p. 674) he goes back upon his former 

 opinion. 



